You're using it on low, that's why. There's a huge difference in performance from low to max effort.

I’m comparing same / similar settings between models. I can’t use high on one and low on others it’s not a fair test.

Not sure why I was downvoted. But seems the downvoter is quick to downvote anything that doesn’t fit the narrative they’re looking for. I’m just reporting my findings.

Low, High and Max, obviously, can't be compared across models. They only mean the model is likely to spend less reasoning effort (~output tokens) with Low than High on the same, *single shot* task.

But even in this very post, you can see that Max was actually cheaper than High.

If you are using API, you should be comparing based on end-to-end cost or speed or whatever blend of those two matches your cost/time budget.

Comparing at similar thinking hasn't much value. You can compare the tiers that have the closest price, that would be more interesting.

Yes, I think a proper comprehensive test would be a better judge of the outcome. I may do round 2 given my first batch of models is already outdated.

I'm not sure it's a fair test either to compare the "low" setting of one model with the "low" setting of another. They're completely different settings that just happen to have the same name.

I did not downvote you. I think it was the way you wrote the comment which made it seem like it couldn't perform in general.

Short via Hyperliquid or some other crypto exchange that tokenizes stock? HL does have a trading pair for MSFT and trades 24/7.

It goes to https://dyne.org/cjit/tutorial.html for me, and very much works as I would expect it to (chromium engine browser, MacOS)

When you're building a social networking site like this, when do you need to start to worry about laws from different states and countries (eg age bans, data export, etc)?

As a guy who used to run game servers[0] and some other small-web stuff, I hate that this is one of the first questions that springs to mind today, but I'd also like to know.

[0]: in ancient times when server meant an actual server

I mean, it's good to be aware from the beginning, even if you don't intend to follow them right away.

But, my rule of thumb is you don't really need to worry about laws from places where they don't have real jurisdiction on you. If they filed suit, would you rather respond or make a note to never visit that place / would you be ok if all your users from there were blocked from contacting you by law?

Site renders extremely poorly on mobile safari that it is completely unreadable.

This is really cool! How are you sandboxing the tiles and allowing limited JS execution?

I'm using JS-Interpreter project: https://github.com/NeilFraser/JS-Interpreter . It's slow, but easy to add and work with.

If I'm on 18.7.1, do I still need to upgrade?

https://www.cvedetails.com/version/2021355/Apple-Iphone-Os-1...

seems to be the same as 18.7.2

https://www.cvedetails.com/version/2037518/Apple-Iphone-Os-1...

Most likely. This is a WebKit issue whose patch is only shipped with iOS 26.2 or iOS 18.7.3 (but that's only available to a subset of iPhone and iPadOS devices).

See: https://support.apple.com/en-us/125885

Paywalled.

You need to update again.

This could be the Next.js motto.

You need to upgrade again, and no the docs aren’t finished (and they won’t be before the new new version).

My Umami stats box got "pwned" about 15 mins after the last CVE was published and I spent an hour or so cleaning up that mess and upgrading everything. Not looking forward to doing it again today.

The second example confuses me. The Person type has isCool: boolean, not an explicit true. How does using satisfies here pass coolPeopleOnly?

You can sorta think of `satisfies Foo` as "the type is exactly the literal value, but also make sure the value could be used in the place of a Foo"